The shadow price of labour

What market reactions show - and what they don't

A point I’ve made a few times is that we need to stop treating market reactions to events as disinterested verdicts. This applies both to bond market reactions to political events and decisions and the comparable reaction of share prices to corporate events and decisions (and obviously there is some overlap).

Market reactions are not an assessment of the general economic (let alone social or environmental) impact of events and decisions. They are an assessment of the impact of those events and decisions on investors. Movements in share price represent the revision in present value of future income streams to investors. They do not represent an assessment of the impact of events and decisions on society. So far so obvious.

Previously I’ve also suggested that sometimes changes in market value might even tell us that an event or decision is good for investors but bad for others. And the scale of the change in market value might give us a sense of how large the underlying redistribution could be.

In this piece I want to float an idea. If we conceive of share price as the present value of future income flows to capital (PVc)¹ is there an equivalent shadow price of the present value of future income flows to labour (PVl)?

A couple of things that have led me to consider this question. The first is the emphasis on ‘firm value’² as metric for valuing corporate decision-making, and even individuals. If a CEO makes a decision and firm value increases this is taken as solid evidence that it was a good decision. And, by extension, if the increase is large this is used as evidence that the CEO is good value for what they are paid. This has always struck me as a very partial picture of what is happening.

Second, I am unconvinced by the argument that one often comes across in the finance sector to focus on ‘growing the pie’ rather than ‘splitting the pie’. This is not because I don’t think win-wins exist. Although win-wins have been oversold in my field in the recent past, clearly mutually beneficial economic activity is not uncommon. But, that said, I believe there are several complicating factors.

First, saying focus on absolute rather than relative shares can be used as intellectual cover for exploitation. Second, ‘growing’ vs ‘splitting’ is not always a choice that is available so calling for a focus on the first can be a bit impractical. Third, relatedly, clearly many decisions straddle the boundary, a decision might increase overall wealth but the extra might accrue to only one party. Fourth, it’s not clear to me that we spend much time trying to measure the impact on the relative pie shares.

The problem is not that the distinction between ‘growing’ and ‘splitting’ the pie is necessarily wrong, but that we lack a way of identifying which is taking place in any given case. In practice, we tend to ask first whether PVc increased (a maths question). Afterwards we ask in general terms if anyone else lost out.

Currently we don’t get an immediate numerical read out on the economic impact (positive or negative) of decisions on workers like we do for investors. So we don’t see the impact on either total value available to the two major stakeholders (let’s stick to a very simplified model of Total = PVc + PVl for now) or the relative change. What we currently lack is a way of analysing corporate decisions that captures both the change in total value and how that value is distributed.

PVc and PVl are not intended to form a precise accounting identity, but rather provide a way to think about different valuations of how corporate decisions affect future income streams from the perspective of different groups. Let’s stick with ‘streams’ as a way of thinking about future income. In a simplified model there are two income streams, valued as PVc and PVl. In this model PVl is the expected value to workers³ of pay, benefits and so on for the workforce in the company under consideration.⁴

Both PVc and PVl are stocks derived from flows. The share price (PVc) is the present value of future income flows to capital, PVl would be the present value of future income flows to labour. There is no double counting here. PVc captures the impact of those flows on investors, while PVl would capture their impact on workers. The same underlying changes can affect both, but they are not the same measure.

Using this model, a corporate event or decision ‘upstream’ could act to increase or decrease the total volume of flows (and thus total value - PVc + PVl), or divert a greater share to one or other channel, or both.

Let’s work the idea through with a simple example. Suppose a firm announces a restructuring that reduces its wage bill by £20m per year. The share price rises, reflecting the higher expected future profits. In present value terms, this might imply an increase in PVc of, say, £200m. So, from the perspective of PVc, the decision is clearly net positive. PVc may incorporate some effects on labour, such as changes in productivity, but only to the extent that they affect expected income to investors.

The same decision clearly reduces the expected future income of workers. If those lost wages and benefits would otherwise have persisted, the present value of that income stream, PVl, may fall by a similar order of magnitude. So the impact on PVl is negative. In that case, what appears as value creation, if considering PVc alone, might be largely a redistribution of future income from labour to capital.

It could even be worse. Suppose investors discount the impact of the reduced costs by reference to reduced productivity. PVc may still increase, but only to £150m rather than £200m, while the hit to PVl remains unchanged. PVc+PVl has the potential to be lower after the decision, even if PVc alone is higher. In that case, a positive share price reaction could coincide with overall value destruction.

But without an estimate of PVl, we cannot tell whether this is a case of overall value creation (or destruction) or simply a reallocation of an existing income stream, or a combination of both. Instead, currently we only see PVc in real time, which is often treated as a proxy for total firm value. In the terms expressed above, that is clearly a partial and one-sided measure.

Currently the market prices the impact on investors of corporate events and decisions immediately. The impact on workers is real, but remains numerically unobserved, in the shadows.

1. If we were to do this properly PVc also needs to include debt, to make it closer to Enterprise Value, but I’m just floating the general idea in this piece.

2. I think it’s quite common to use just raw market cap (PVc) not even Enterprise Value.

3. This is important. PVl would not be a measure of ‘human capital’, which is an attempt to capture the value of labour to the firm. It would be a measure of the value of labour to the workers themselves.

4. This is a firm-level measure. Workers might find more favourable employment elsewhere, but the aim here is to capture how a given decision affects expected income streams for a workforce within a specific firm, in the same way that the share price reflects investor value in a specific company. I’m not advocating a labour equivalent of the All-Share just yet.